Transantarctic Mountains, Antarctica

Home , Dominion Range, Nothofagus, Queen Alexandra Range

Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 www.elsevier.com/locate/palaeo

Neogene vegetation of the Meyer Desert Formation (Sirius Group) Transantarctic Mountains, Antarctica

Allan C. Ashwortha,*, David J. Cantrillb

aDepartment of Geosciences, North Dakota State University, Fargo, ND 58105-5517, USA bDepartment of Palaeobotany, Swedish Museum of Natural History, Box 50007, SE-104 05, Stockholm, Sweden Received 5 November 2003; received in revised form 28 May 2004; accepted 2 July 2004

Abstract

A tundra vegetation consisting of at least 18 plant species is described from the Meyer Desert Formation which outcrops along the Beardmore Glacier in the Transantarctic Mountains, about 500 km from the South Pole. The fossils include pollen, seeds, fruits, flowers, leaves, wood, and in situ plants, of which wood and leaves of Nothofagus and a pollen assemblage had been previously reported. The plants include a cryptogamic flora of mosses and liverworts, conifers, and angiosperms in the families Gramineae, Cyperaceae, Nothofagaceae, Ranunculaceae, Hippuridaceae, ?Caryophyllaceae, and ?Chenopodiaceae or ?Myrtaceae. The plants grew in a weakly-developed soil formed within a complex periglacial environment that included moraines, glacial outwash streams, well-drained gravel ridges, and poorly drained depressions in which peat and marl were being deposited. D 2004 Elsevier B.V. All rights reserved.

Keywords: Antarctica; Neogene; Nothofagus; Palaeoecology; Biodiversity; Palaeoclimate

1. Introduction yophyllaceae). Both species occur in the area from the islands of the Scotia Ridge, along the west coast of the The Antarctic flora today is dominated by crypto- Antarctic Peninsula south to the current southernmost gams, most of which only grow in the most sheltered site on Alamode Island, Terra Firma Islands at coastal locations north of 658S. Only two species of 68842VS. On the eastern side of the Antarctic vascular plants are known, a grass Deschampsia Peninsula, adjoining the frozen Weddell Sea, vascular antarctica Desv. (Poaceae) and a herbaceous caro- plants extend only as far south as 658 (Kappen and phyllid, Colobanthus quitensis (Kunth.) Bartl (Car- Schroeter, 2002). Both D. antarctica and C. quitensis are considered to have colonized the Antarctic Peninsula during the Holocene. Convey (1996) felt * Corresponding author. Fax: +1 701 231 7149. that the main reason why only two species had E-mail address: [email protected] successfully invaded was because of the geographic (A.C. Ashworth). isolation of Antarctica. The asymmetry in the geo-

0031-0182/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2004.07.002 66 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 graphic ranges of the species on the Antarctic each of a freshwater bivalve and a lymnaeid gastropod Peninsula, however, suggests that climatic factors (Ashworth and Preece, 2003). The new fossil plant play an important role in their distribution. Support materials reported here include moss stems and for the importance of climate in limiting distribution leaves, in situ cushion plants, and fruits, stems and comes from the failure of experiments to transplant flowers of several species of vascular plants. The the species further south (Lewis Smith, 1994). discoveries indicate a more diverse Neogene flora The simple vegetational communities and the low than was previously known and, together with the plant diversity seen today contrast with the rich fossil previously discovered fossils, enable a more complete record. Diverse forests of conifers (podocarps and interpretation of the Neogene vegetation and land- araucarians) and angiosperms (Cunoniaceae, Notho- scape in the Transantarctic Mountains. fagaceae, Proteaceae, Lauraceae) inhabited the continent during the Late Cretaceous and Paleogene Periods (Dettmann and Thomson, 1987; Askin, 2. Stratigraphy and age 1992; Cantrill and Poole, 2002). The geographic isolation associated with the breakup of Gondwana, The Meyer Desert Formation (MDF) is the name and the development of the Antarctic ice sheets, given to the mostly nonmarine glacigenic deposits resulted in radical changes to the vegetation. How- forming the upper part of the Sirius Group, formerly ever, the nature and the timing of extinctions caused the Sirius Formation, in the Meyer Desert and by climatic cooling is still being debated largely due Dominion Range region of the Transantarctic Moun- to the paucity of Neogene fossil sites (Cantrill, 2001). tains (Mercer, 1972; McKelvey et al., 1991; Webb et Mercer (1972, p. 432) was the first to suggest that al., 1996). The deposits are gently inclined to the an examination of the contact between the Sirius north and are fractured by numerous normal faults Formation, a formation he named, and the underlying with throws of up to 300 m (McKelvey et al., 1991). bedrock in the Oliver Bluffs, Dominion Range, might The MDF occurs in discrete, terraced sedimentary result in the discovery of d...a palaeosol, or traces of packages, with a combined thickness of about 185 m. vegetation.T His intuition proved to be correct. Each of the packages is unconformable on the Nothofagus wood was described by Carlquist (1987) bedrock of a multi-tiered, glacially-scoured erosion and Webb et al. (1987) and pollen by Askin and surface, representing several distinct episodes of uplift Markgraf (1986). Further descriptions of the wood, and erosion followed by deposition. pollen, and well-preserved leaves of a new species of The fossils are from beds within the topographically Nothofagus, Nothofagus beardmorensis, were made lowest and stratigraphically younger part of the by Hill and Truswell (1993), Webb and Harwood formation exposed in the Oliver Bluffs, on the eastern (1993), Hill et al. (1996), and Francis and Hill (1996). flank of the upper valley of the Beardmore Glacier at Today, the climate of the Oliver Bluffs is too cold and latitude 85807VS, longitude 166835VE. The Oliver dry to support vascular plant or cryptogram growth. Bluffs, at an elevation of 1760 m above sea level, are The reports of Neogene fossil wood, leaves, and located about 170 km south of the confluence of the pollen of Nothofagus from a site so far south are one Beardmore Glacier with the Ross Ice Shelf and about of the most important palaeobotanical discoveries to 500 km north of the South Pole (Fig. 1). The new fossils have been made in Antarctica. The discoveries are from a horizon about 20 m above the base of an provided the basis for palaeoclimatic interpretations erosional spur located near the northern end of the that became significant in an ongoing debate about the bluffs (Fig. 2A). The section is about 900 m north of the relative stability of the Antarctic ice sheets. one from which leaves of Nothofagus beardmorensis An expedition to the Oliver Bluffs during the 1995 were described (Hill et al., 1996). The section consists austral summer resulted in the discovery of new fossil of 85.5 m of interbedded diamictites, coarse cross- materials. These included two species of curculionid bedded sandstones, and wood-bearing siltstones and beetles (Ashworth et al., 1997; Ashworth and sandstones. On the scree slope immediately north of Kuschel, 2003), one species of a cyclorraphid fly the spur, we discovered a thin lithified peat interbedded (Ashworth and Thompson, 2003), and one species between thin marlstone deposits (Fig. 2F), at a similar A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 67

Fig. 1. Locality map of the fossil site at the Oliver Bluffs on the Beardmore Glacier in the Meyer Desert region of the Dominion Range, Transantarctic Mountains. stratigraphic level to the wood-bearing siltstones and Group in the Beardmore region, outcrops about 90 km sandstones. The horizon is also at about the same level down-glacier from the Oliver Bluffs. Webb et al. (1996) as a palaeosol identified as the lower of two Viento interpreted the non-stratified and stratified drift with types by Retallack et al. (2001, Fig. 5). McKelvey et al. dropstones as glaciomarine sediments deposited by (1991) interpreted the diamictites as lodgement tills grounded and floating ice in a wide fjord. The deposited by the ancestral Beardmore Glacier during a Cloudmaker Formation is overlain by the MDF series of advances and retreats; they interpreted the implying that the fjord infilled with sediment either fossiliferous deposits to have accumulated during an by subsidence and progradation or by changes in base interglaciation. level. Webb et al. (1996) also proposed that subsequent Agglutinated foraminifera and other marine micro- to deposition, the Sirius Group deposits were uplifted fossils occur in the tillites at the southern end of the by about 1300 m. bluffs below the horizon containing leaves of Notho- A biostratigraphic age of less than 3.8 Ma has been fagus beardmorensis. Webb et al. (1996) correlated the assigned to the MDF based on the occurrence in the foraminifera assemblage with a similar assemblage Oliver Bluffs glacigenic sediments of reworked from the upper part of the Cloudmaker Formation. The marine diatoms (Harwood, 1986; Webb et al., 1996). Cloudmaker Formation, the lower part of the Sirius The same diatom taxa occur in situ with tephra in the

A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 69

CIROS-2 core from the Ross Sea with 40Ar/39Ar ages abundant achenes (all types), along with fragments of of 2.77 Ma (Barrett et al., 1992). Nothofagus pollen of mosses, wood, insects, and ostracods. the same type that occurs in the MDF has also been All fossils were mounted on micropalaeontolog- reported from DSDP Site 274 in the Ross Sea with a ical slides for preservation and are part of the biostratigraphic age of about 3 Ma (Fleming and collection of the Quaternary Entomology Laboratory Barron, 1996). The pollen showed no signs of having at North Dakota State University. Statistical analysis been reworked and Fleming and Barron considered of the achenes involved accurately drawing outlines the most likely source would have been from of every specimen at a magnification of about 25Â Nothofagus growingontheRossSeacoast.A using a camera lucida attached to a binocular Pliocene age is also supported by the degree of microscope. The images were then scanned and the weathering and development of palaeosols in the area and lengths of major and minor axes were MDF (Retallack et al., 2001). The palaeosol closest to determined (in pixels) using a Fourier-based method the fossil horizon was assigned an age of between 1.3 described by Cheng (2003). The measurements were and 4.1 Ma (Retallack et al., 2001). then converted from pixels to millimeters to plot the The biostratigraphic age assignment depends on graphs (Fig. 7). the diatoms being incorporated in the glacial deposits directly from the reworking of older marine sediments. That interpretation has been challenged in 4. Biodiversity several studies in which it is considered more probable that the diatoms were wind-deposited 4.1. Cryptogams (Burckle and Potter, 1996; Burckle et al., 1996; Kellogg and Kellogg, 1996; Stroeven et al., 1996; Fossil moss cushions were reported to occur in the Barrett et al., 1997) or from the fall-out of sediments MDF at the Oliver Bluffs (Hill et al., 1996), but no blasted from the deep ocean floor by the Eltanin descriptions of them were made. In our samples, asteroid (Gersonde et al., 1997). In either case, the vegetative axes that lack vascular tissue are common MDF is considered to be older than the diatoms. and these are interpreted as the remains of mosses. In all, more than 500 fragments were recovered. The majority, however, are small (less than a few milli- 3. Field and laboratory methods meters long), unbranched, and often leafless, although the leaf bases may be present. The fragmentary nature Approximately 400 kg of samples were collected of the axes mostly precludes identification to the from the different lithologies and shipped to the generic or family level. Where whole leaves were laboratory for preparation. All samples were disaggre- preserved, leaf arrangement, morphology, and cell gated first by soaking in water to soften the calcite structure were used for identification. We identified at cement and then by wet sieving through a 300-Am least five different types of mosses but we believe that sieve. The fossils were then picked from the grain represents a minimum number of species. residue under a binocular microscope. The samples that Moss type 1 comprises axes with distinctly preserved the best cellular detail were obtained from distichously arranged leaves and is represented by about 50-cm depth, indicating that oxidation of organic 15 specimens (Fig. 3A). Leaves are strongly recurved material is occurring at the surface of the outcrop. and a central nerve extends from base to apex (not Lithologies varied with respect to the yield and the excurrent). Leaf cells of the lamina are small and composition of fossils. The marl contained fragments rectangular, ranging from 10 to 16 Am long by 10 to of mosses along with freshwater molluscs (shells of the 16 Amwide(Fig. 3B). The upper leaf surface is bivalve Pisidium and a species of lymnaeid gastropod; smooth and the lower surface is strongly papillate Ashworth and Preece, 2003), a single fish tooth, "and (Fig. 3D), but less so over the midregion of the leaf. rare achenes (Ranunculus type). The peat yielded moss Cells over the central nerve are c. 25–45 Am long by and shell fragments and some achenes (Ranunculus 9–11 Am wide. The strongly papillate leaves suggest type), while the siltstones yielded cushion plants, affinities with the Pottiaceae. 70 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

Fig. 3. Scanning electron micrographs of cryptogamic fossils. A, B, and D: Moss type 1. (A) Overview of habit with strongly recurved leaves. (B) Cell arrangement on upper leaf surface. (D) Lower leaf surface adjacent to midrib. Note the prominent papillae. C and E: Moss type 2. (C) Habit of axes with small widely spaced appressed leaves. (E) Close up of individual appressed leaf. F: Moss type 3. G and H: Spore bearing body. (G) Small ovate body with medial ridge that apparently bifurcates. (H) Small smooth walled 20–30 Am spores with trilete marks. Scale bars A and G represent 1 mm; B, C, E, and F represent 200 Am; D and H represent 50 Am. A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 71

Moss type 2 is represented by two strands, the longest c. 5 mm long (Fig. 3C). This axis is covered by widely spaced and appressed linear lanceolate leaves, up to 0.4 mm long. The leaf margin is entire, and the leaf apex is narrowly acute and tapered (Fig. 3E). Cells of the lamina are composed of small rectangular cells 10–18 Am long by 8–10 Am wide. Moss type 3 appears to have four ranks of leaves that are strongly imbricate. The leaves are ovate to lanceolate, 0.3–0.6 mm long by 0.2–0.4 mm wide; leaf surfaces are smooth, margins entire, apex rounded (Fig. 3F). Moss type 4 is by far the most common component of the moss flora represented by at least 150 specimens. The majority of the axes are missing leaves but leaf bases are intact. The axes are generally unbranched, but occasionally, there are rare lateral branches. Leaves are arranged in three or four ranks, lanceolate, and lacking a central nerve (Fig. 4A). Leaf margins are entire and leaf apices are acute (Fig. 4A). Leaves are inserted at greater than 458. Leaf cells are uniformly elongate with length–width ratios between 8 and 10. Cells range from 60 to 100 Am long by 7 to 10 Am wide. No specialized cells around the leaf bases were observed. Moss type 5 is represented by four isolated leaves. However, the distinctive nature of the leaf material indicates that this is a different species. The leaves are small (0.65–0.8 mm long by 0.4 mm wide), ovate with a contracted base, and a prominent nerve that extends to the leaf apex (Fig. 4B). The nerve is compound comprising three to four rows of cells. Fig. 4. Light micrographs of plant fossils. (A) Moss type 4 with simple lanceolate leaves. (B) Moss type 5 with small ovate leaf that Cells of the nerve are elongate, 80–100 Am long by have a prominent central midrib and isodiametric cells. (C) Portion 10–12 Am wide. Cells of the leaf lamina are of ?grass blade with a central vascular strand surrounded by cells rectangular, rarely polygonal, and 18–32 Amin filled with brownish material. Scale bars A–C represent 0.5 mm. diameter (Fig. 4B). No reproductive mesofossils of cryptogams were A single fossil, 1.5 mm long by 1.0 mm wide, identified, but spores have been reported from the obovate in shape, may represent a part of a lower plant deposits. Smooth, thick-walled bodies, or resting (Fig. 3G). The base is tapered and terminated in a spores, point to the presence of algal and fungal semicircular structure that superficially resembles a groups (Askin and Markgraf, 1986; Hill and Truswell, leaf attachment point. A single vein-like ridge arises at 1993). A low diversity assemblage of bryophytes is the base and branches once. The surface is smooth, also indicated by three Coptospora species and but cell structure is visible in some places and Marchantiaceae spores (Askin and Raine, 2000). comprises small rectangular cells. The ridge-like vein Palynological examination of the marls and peat is fractured, and the interior is packed with small, (Askin and Ashworth, 1998) yielded similar low triangular, smooth-walled trilete spores (Fig. 3H), diversity assemblages. ranging from 15 to 20 Am in diameter. 72 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

4.2. Conifers bifurcating branching habit towards the outside of the mound (Fig. 2E). The axes range from 2 to 8 mm in Conifers have been recorded in the pollen diameter. The density of the outer surface and the assemblages by the presence of rare bisaccate packing of the axes indicate a compact habit rather pollen grains initially identified as ?Dacrydium than an open shrubby habit. The axes are covered (Askin and Markgraf, 1986), but more recently basally in leaves. The leaves, however, are not well- referred to Podocarpidites sp. b (Askin and Raine, preserved except for the outer ones which are linear to 2000). The only other reference to conifers is that linear lanceolate (Fig. 5A). The woody texture of fossil wood (Carlquist, personal communication, suggests that the mounds were formed by vascular in Askin and Markgraf, 1986), but a later study plants with a cushion-like habit. This is confirmed by (Carlquist, 1987) stated all of the wood collected vascular tissue seen in the leaf and the recognition of belonged to Nothofagus. Francis and Hill (1996) stomata on the leaf surface (Fig. 5B). A cushion-like reported only Nothofagus wood, and all of the habit is found in many southern hemisphere angio- wood from our samples is angiospermous. sperm families such as Caryophyllaceae (e.g. Colo- banthus), Apiaceae (e.g. Azorella), Umbelliferae (e.g. 4.3. Angiosperms Bolax), Donatiaceae (e.g. Donatia), and Asteraceae (e.g. Haastia, Raoulia). Other vegetative material Flowering plants are represented by vegetative includes a long, linear leaf fragment that is 16.5 mm material including leaves and wood and reproductive long by 0.5 mm wide that may be part of a grass structures including pollen, seeds and fruits, and blade. Internally, the leaf has a central vascular strand flower parts. Much has been written about the surrounded by distinctive brown-coloured cells (Fig. spectacularly preserved Nothofagus leaves (Hill and 4C). The brown colouration is confined to small Truswell, 1993; Hill et al., 1996) and wood (Webb regions within the cells, and these are interpreted to be and Harwood, 1987; Carlquist, 1987; Francis and Hill, the remains of chloroplasts. The position of the cells is 1996) from the MDF and is not repeated here. The similar to bundle sheath cells observed in grasses. pollen record is dominated by single type of Notho- Reproductive materials support the presence of fagidites that was compared to Nothofagidites lach- more vascular plant diversity. A single flower was maniae (Askin and Markgraf, 1986; Hill and recovered from the organic residues. The flower is Truswell, 1993) and placed in the subgenus Fusco- not well-preserved but terminates a robust pedicel spora (Hill and Truswell, 1993). It should be noted (Fig. 4E–G). Unfortunately, the abrasion of the that N. lachmaniae is placed in the subgenus material has resulted in the loss of the outer whorls Nothofagus by Dettmann et al. (1990). of the calyx, but scars show the position of parts. Although Nothofagus pollen dominates the The outer whorl of parts surrounds a central cup that assemblages, other angiosperm diversity is recognized appears to be five partite (Fig. 5G). Details of the by the presence of rare tricolpate grains (Tricolpites anthers and ovary are not preserved, but the flower sp. 2 of Truswell, 1983; Tricolpites sp. a of Raine, appears to have a superior ovary. The present lack of 1998; Askin and Raine, 2000) that is thought to be detail of floral number and arrangement precludes related to Polygonaceae or Laminaceae (Askin and systematic placement. Markgraf, 1986; Hill and Truswell, 1993; Raine, A poorly preserved structure thought to represent 1998). another flower type was also identified (Fig. 5C and Vegetative remains indicate further diversity to this D). This structure appears to have two outer bracts sparse pollen record. Within the fossiliferous horizon subtending a complex of perianth parts. By far, the at the Oliver Bluffs, mounds up to 30 cm in basal most interesting reproductive organs recovered from diameter and 16 cm in height are encountered (Fig. 2C the MDF deposits are 366 specimens of small seeds and E). Although moss cushions have been recorded and fruits ranging from 0.4 to 2.0 mm long (Fig. 6). (Hill et al., 1996), this material differs substantially. There are five distinctive morphologies, but the Internally, the mounds are fibrous in appearance and majority (98.7%) probably represent one genus and composed of robust woody axes radiating with a possibly one species. A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 73

Fig. 5. Mesofossils of cushion plants and flowers. A and B: Cushion Plant. (A) Portion of axis with imbricate linear leaves. (B) Fractured surface of the leaf revealing poorly preserved stomata. C and D: Small inflorescence. (C) Lateral view of possible bipartite inflorescence. (D) Apical view of same structure. E–G: Flower. (E) Overview of whole axes terminating in flower. (F) Lateral view showing perianth parts. (G) Apical view illustrating five partite arrangement of perianth. Scale bar A represents 5 mm; B represents 50 Am; C–G represents 1 mm. 74 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

Fig. 6. Dispersed seed and fruit flora. A, C, and D: Type 1 fossil achenes. (A) More complete specimen. (C) Eroded specimens. (D) Laterally compressed specimen revealing ventral suture. (B), Modern achene of Ranuculus peduncularis Sm. from Tierra del Fuego for comparison with the fossils. E: Type 2 fruit, a barrel-shaped seed identified as Hippuris. F: Type 6—?seed. G: Type 5, small reniform seed. H: Type 3 fruit identified as Cyperaceae. I: Type 4 fruit identified as Cyperaceae. Scale bars A–C represent 2 mm; D–F and I represent 1 mm; G and H represent 0.5 mm. A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 75

Type 1 are achenes that are broadly elliptical to obovate, with a more or less straight to slightly convex ventral side, and a distinctly convex dorsal side (Fig. 6A, C, and D). The specimens, which are laterally compressed, range in size from 1.1 to 1.7 long by 0.6 to 1.3 wide. The bases are narrow, often with distinct oblique attachment points (Fig. 6A). On a few specimens, the apex is prolonged into a short beak, but this feature has been abraded in the majority of specimens. Where the outer wall has not been abraded, the achenes appear to be smooth. The cells exposed in the abraded walls are equidimensional, randomly oriented, and 20Â20 Am in size. Occasionally, two distinct veins are present that are characterized by longitudinally orientated cells. These arise in the attachment point and extend for up to three quarters of the achene length. The distinctive shape of the better-preserved achenes is most similar to that seen in extant Ranunculus species (buttercups) (Fig. 6B) with which we compare it here. The lengths of major and minor axes were plotted against area for 328 of the better preserved specimens. Significant regression analyses 2 Fig. 7. Regression analyses of area in relation to (A) minor axes r and p values indicate that the morphological 2 2 (r =0.745, pV0.001) and to (B) major axes (r =0.537, pV0.001) for variation is continuous (Fig. 7). The most parsimo- 328 fossil achenes of Ranunculus. nious interpretation of the data is that they represent a single species, with the variation attributable to a combination of natural and taphonomic causes. Some Type 3 is a fruit represented by a single distinctive variation in morphology results from the position in specimen that is 1 mm long by 0.7 mm in diameter, which the achenes developed on the inflorescence broadly fusiform, with a distinct small basal stalk, and head. Other variation resulted from abrasion of the a tapering apex (Fig. 6H). The specimen has the achenes before burial and later crushing and distortion morphology of a Cyperaceae fruit (sedges). in the sediment. Type 4 is a flattened fruit that is elliptical to Four additional types of seeds or fruits are also ovate in outline, 1.75 mm long by 1.25 mm wide present but are represented by one or a few specimens. (Fig. 6I). It has a distinct contracted basal attach- Type 2 is a fruit represented by a single distinctive ment and apically appears to terminate in a specimen that is 1.74 mm long by 0.98 mm wide, contracted and possible mucronate style. The outer oblong, spherical, with a distinct small basal stalk and surface has a faint ornament of raised cell walls. a round opening in the apex (Fig. 6E). The lack of The specimen has the morphology of a Cyperaceae material makes comparisons difficult, but similarities fruit but of a different species than represented by with extant Hippuris (mare’s tails) are remarkable. Type 3. Hippuris is a widespread, perennial herb of shallow Type 5 is a small reniform seed, campylotro- water and mud flats, occurring from high arctic to pous, 0.75 mm long by 0.5 mm in diameter, and temperate latitudes in North America and Eurasia, but slightly asymmetrical (Fig. 6G). Cells on the outer also occurring in Australia and southern South surface are transversely orientated and the anticlinal America where it is native (Cronquist, 1981; Moore, walls are smooth. The placentation seen in this type 1983). The bipolar distribution is believed to result is rare in angiosperms but is widespread in the from dispersal of fruits or rhizomes by migratory birds caryophyllids including families such as Chenopo- (e.g. Arctic terns). diaceae and also in families such as the Myrtaceae. 76 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

It is interesting to note that Wilson et al. (1998) the beds immediately below the fossiliferous beds is a recorded a Chenopodiaceae pollen grain from Sirius strong indication that deposition occurred close to sea Group strata in the Reedy Valley. level. Furthermore, the occurrence of glaciomarine In addition to the five types of achenes and fruits, deposits of the Cloudmaker Formation down valley there are two additional specimens that may represent indicates marine inundation and the existence at times a sixth type. The specimens are spherical and broadly of a fjord as far inland as the Oliver Bluffs. ovate in shape, 0.3–0.5 mm in diameter (Fig. 6F). Postdepositional uplift suggests that landscape relief at the time of deposition of the MDF was more than a thousand meters lower than today. The stratigraphy at 5. Sedimentary environments, the physical the Oliver Bluffs of multiple lodgement tillites, landscape, and the extent of the East Antarctic Ice interbedded with wood-bearing sandstones and silt- Sheet stones, is evidence of a dynamic ice margin during which the glacier advanced and retreated on several The Sirius Group, formerly the Sirius Formation, is occasions. The occurrence of palaeosols at several patchily distributed throughout the Transantarctic horizons indicates episodes of relative stability during Mountains (TAM) at elevations ranging from 1750 which soils formed (Retallack et al., 2001). The to 3000 m (Mayewski and Goldthwaite, 1985). lithofacies variations over short distances within the Deposition of the tillites, fluvioglacial, glaciolacus- fossiliferous bed suggest a generally unstable sedi- trine, and glaciomarine deposits clearly reflect a mentary environment. We envisage a landscape of warmer and wetter climatic regime than exists in low-relief moraines and a wide, braided outwash plain Antarctica today. The relationship of the deposits, that existed between the East Antarctic Ice Sheet, the however, to the extent and the dynamics of the East glacier margin, and the fjord. The moraines provided Antarctic Ice Sheet (EAIS) is contentious. Passchier more stable surfaces for plant colonization, but plants (2001) summarized various hypotheses that have been would also have been able to have colonized the proposed. The first is that deposition was from gravel bars and ridges between the channels of the glaciers that flowed from ice caps developed on braided meltwater streams. Abandoned channels on plateaus in the TAM during the early stages of uplift the margins of the active system would have provided and prior to the formation of the EAIS (Mercer, 1972). locations in which thin peats and marls accumulated. The second is that the high elevation Sirius deposits in In many respects, the appearance of the landscape the Dry Valleys were deposited by alpine glaciers and would have been similar to that of the fjords of not by outlet glaciers of the EAIS, although it was in western Greenland today (Fig. 2B). existence (Stroeven and Prentice, 1997). The third is that the deposits represent the overriding of a preexisting TAM by a greatly expanded EAIS during 6. Diversity and plant palaeoecology an episode of major continental glaciation (Mayewski, 1975; Webb et al., 1984; Denton et al., 1984). The Antarctica has been essentially in a polar position fourth is that the higher elevation Sirius deposits are since the Early Cretaceous (Lawver et al., 1992) and associated with an EAIS that overrode the TAM and at 858S receives no sunlight from the middle of March that the younger deposits are those associated with until the end of September. The total amount of solar valley glaciers that incised the TAM as a result of radiation received today is about 2250 W/m2 and the uplift (Brady and McKelvey, 1979, 1983; Barrett and number varies by a small amount because of changes Powell, 1982; Webb et al., 1996; Van der Wateren et in the eccentricity of the earth’s orbit. The amount of al., 1999; Passchier, 2001; Hambrey et al., 2003). radiation received is about 42% of that received over For the deposition of the MDF, we prefer the last Tierra del Fuego at 558S. The absence of solar heating mentioned interpretation. We believe that the EAIS for about 5.5 months, even though there would have was already in existence, and the ancestral Beardmore been additional heat from the ocean, would have Glacier was already within a valley incised into the resulted in a low mean annual temperature. We expect TAM. The occurrence of agglutinated foraminifera in that the indirect effects would have been low A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 77 atmospheric moisture content, low annual precipita- palaeosol (Viento-type) exhibiting weakly-developed tion, a short growing season, slow rates of chemical structures characteristic of a permafrost landscape weathering and low nutrient availability, especially (Retallack et al., 2001). The asymmetric growth rings nitrogen and phosphorus, and permafrost with of fossil Nothofagus wood indicate a low sprawling restricted drainage and poor aeration. Furthermore, habit for these shrubs with some evidence for wind strong adiabatic winds from the glacier, winter snow abrasion (Francis and Hill, 1996). The exceedingly cover, summer snow and ice storms, and frosts narrow growth rings (Francis and Hill, 1996) further throughout the growing season would have contrib- indicate an environment in which soil-forming and uted to a severe environment for plant growth. biotic processes were slow. The growing season, sandwiched between early Within the fossil horizon are in situ cushions of summer snow melt and killing autumnal frosts, would vascular plants and mosses. The woody tissue of the be expected to be between about 6 and 12 weeks long. vascular plants forms cushions that are 30 cm in Above-ground vegetative growth would occur mostly diameter at the base and 16 cm high. Cushion plants immediately after snow melt. Root growth would be in tundra habitats are characteristic of wind-exposed delayed by the frozen soil but would continue after sites with well-drained soils (Bliss, 1997). The speci- killing frosts. Photosynthesis would continue through- mens evidently were sufficiently well-rooted to be out the 24-h photoperiod enabling plants to grow buried in their growth positions by outwash deposits. rapidly. Flowering, reproduction, and seed production Leaves from Nothofagus beardmorensis form a would probably only occur in the summers with the dense leaf mat that probably represent a deciduous most favourable climatic conditions. leaf fall and accumulation in a shallow pool of water Plant species in the High Arctic are so well- (Webb and Harwood, 1993). In polar latitudes, adapted to low temperature that they have evolved shrubby deciduous taxa are restricted to the warmest physiologies and morphologies that enable them to and most protected habitats (Bliss, 1997). Evidence survive winter desiccation and where special adapta- from growth ring anatomy of the fossil wood indicates tions for photosynthesis, respiration, and absorption of that the shrubs had a low prostrate growth form nutrients enable them grow at similar rates to those in (Francis and Hill, 1996). We consider this a more more temperate latitudes (Chapin and Shaver, 1985). realistic interpretation than the patch of upright shrubs The MDF plants also had to be highly-adapted to exist illustrated by Webb and Harwood (1993). in the reduced photoperiod and low temperatures at Peat lenses, between 1 and 10 cm in thickness, lat. 858S. occur within the fossiliferous horizon. At one locality, The plants of the MDF flora, based on compiling a thin peat about 5 cm in thickness is underlain and various lines of fossil evidence, include at least three overlain by marl deposits about 2 cm in thickness; the types of liverworts, five species of mosses, a species marls contain abundant broken shells of a species of of conifer (Podocarpaceae), and at least seven species lymnaeid gastropod and a species of Pisidium,a of vascular plants, including representatives from the fingernail clam (Ashworth and Preece, 2003). The grasses (Gramineae), sedges (Cyperaceae), southern peat and marl beds were deformed during a glacial beeches (Nothofagaceae), buttercups (Ranuncula- advance and were squeezed up between the boulders ceae), mare’s tails (Hippuridaceae), and ?Caryophyl- of a lodgement till. The peats contain only occasion- laceae, and ?Chenopodiaceae or Myrtaceae. The ally recognizable moss stems suggesting that they are number of taxa is higher than has been reported either strongly humified moss bank or mire deposits, previously from the MDF but still at the low level of or that they are microbial mats formed by cyanobac- diversity that would be expected for a tundra flora teria and algae. growing at lat. 858S. The variety of growth forms in the MDF fossil We surmise that the first colonizers on the land- assemblage represents a typical tundra assemblage in scape at the head of the Beardmore fjord would have which there is selection in the harsh environment for been cushion-forming species of mosses that trap phenological plasticity. The growth forms of dwarf organic matter and contribute to soil formation (Lewis and prostrate shrubs, cushion, and graminioid would Smith, 1994). The fossils occur in association with a be adaptive to the different microtopographic sites. 78 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

The cushion forms would mostly be morphological phers gained the upper hand, although it has always adaptations to the most exposed windswept locations. been understood that the natural terrestrial biota of We envisage a tundra environment in which there was remote, geologically young, volcanic subantarctic a mosaic of well-drained and poorly-drained micro- islands must have reached their destinations either sites in which nutrient availability would have been by transport on ocean currents or by birds. Molecular patchily distributed. Mean summer temperatures, genetic studies of plants, including species of Ranun- based on the minimal thermal requirements for culaceae, suggest that many plant relationships, Nothofagus, listroderine weevils, and freshwater formerly considered to be ancient, may be consid- molluscs, are estimated to have been about 4 to 5 erably younger, and that their current distributions 8C for at least two summer months (Francis and Hill, result from dispersal and not vicariance (Lockhart et 1996; Hill and Jordan, 1996; Ashworth and Kuschel, al., 2001; Winkworth et al., 2002). To the vicariance 2003; Ashworth and Preece, 2003). biogeographers, Antarctica is viewed as an evolu- The fossils we have identified are from deposits tionary centre and, to the dispersalists, as a possible that are interbedded between lodgement tills. The stepping stone between South America and Australia organisms colonized the head of the Beardmore fjord and New Zealand. as the glacier retreated. Those communities were later Potentially, the MDF fossil assemblage could destroyed by glacial advance. The palaeosol associ- represent either a relictual Antarctic biota or one that ated with the horizon is just one of several that have dispersed to Antarctica during warm intervals during been identified in the Meyer Desert Formation the Neogene. Francis and Hill (1996) discussed this (Retallack et al., 2001). Palaeosols with abundant question with respect to Nothofagus beardmorensis. root casts and fossil wood higher in the Oliver Bluffs They concluded that because of the dispersal mech- stratigraphic section indicate multiple phases of anisms available to Nothofagus, it was much more recolonization of the landscape associated with glacial probable that N. beardmorensis was a descendant retreat. Fossil wood, presumably Nothofagus, and from a long lineage of Antarctic species. We cannot moss stems reported from near the base of the MDF really be sure, however, that all the plants in the MDF on the Cloudmaker about 90 km north of the Oliver are relictual (e.g. Ranunculaceae, Hippuridaceae). Bluffs, suggests that a tundra vegetation had a much Molecular genetic studies of a subset of alpine New wider distribution in the Beardmore region (Webb et Zealand Ranunculus species indicate a Pliocene al., 1996). Reports of wood, woody tissue, or moss radiation and subsequent dispersal to Australia and stems from Sirius Group strata, or their stratigraphic the New Zealand subantarctic islands (Lockhart et al., equivalents, at localities separated by hundreds of 2001; Winkworth et al., 2002). This indicates the kilometers in the TAM suggest that tundra may have possibility of long distance dispersal to Antarctica occurred in coastal locations in East Antarctica during the Neogene. However, it does not preclude the extending inland up fjords during warmer intervals, chance that these plants may also be a relictual possibly from the Oligocene to the Pliocene (Wilson component of the Cretaceous and early Tertiary et al., 2002; Hambrey et al., 2003; Marchant, personal Antarctic biota. communication, Marchant et al., 2002). The identification of a Hippuris-like fruit in the MDF is biogeographically fascinating; the implication is that the existing bipolar distribution of Hippuris 7. Biogeography vulgaris L (Water Mare’s Tail) between the northern and southern hemispheres is ancient. We imagine that Ever since J.D. Hooker (1853), biogeographers melting of the EAIS during the Neogene permitted the have debated the relationships between organisms on ancestors of existing bird species to extend their long- the widely separated continents of the southern distance migrations into the fjords deep in the interior hemisphere. Similarities are explained either by of Antarctica carrying with them plants and inverte- reference to ancient Gondwana relationships or by brate animals. more recent dispersal. With the development of the The isolation of Antarctica as a result of plate plate tectonics hypothesis, the vicariance biogeogra- tectonic processes undoubtedly played an important A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 79 role in the evolution of the earth’s climate system. A Oligocene vegetation was uncertain due to poor age general decline in global temperatures through the control and difficulty in interpretation. Tertiary is overprinted by rapid cooling events (e.g. The drilling of the CIROS-1 borehole, with better 33.5 Ma) as indicated by shifts in oxygen isotopes age correlations, provided solid evidence for the (e.g. Zachos et al., 2001; DeConto and Pollard, 2003), presence of a depauperate Oligocene vegetation and these signal the initial formation of Antarctic ice represented by a Nothofagus leaf (Hill, 1989) and a and a subsequent rapid expansion of Antarctic ice sparse pollen record (Mildenhall, 1989). A decrease in sheets. The cooling of Antarctica and the associated species richness is noted higher in the stratigraphic reorganization in oceanic and atmospheric circulation section in both the CIROS-1 and MSSTS-1 boreholes had a profound effect on the vegetation of the (Truswell, 1990), pointing to a progressive decline in continent. Prior to the Oligocene, the floras of diversity and vegetation. Similar patterns have Antarctica were relatively diverse, although it is clear recently been confirmed by the Cape Roberts Drilling that cool temperate elements such as Nothofagus Program where a single Nothofagus leaf occurs in the become increasingly important during the Eocene in Oligocene (Cantrill, 2001), and the palynological regions such as the Ross Sea (Pole et al., 2000) and assemblages show a species-poor flora that decreases the Antarctic Peninsula (Poole, 2002;Pooleand in abundance and diversity from the Oligocene Cantrill, submitted for publication). Following a through the early Miocene (Raine, 1998; Raine and major cooling event around the Eocene/Oligocene Askin, 2001). The palynological record clearly boundary, there is a marked increase in glacially- indicates that a tundra vegetation, containing vascular derived sediments, and the record of fossil vegetation plant species, existed in the Ross Sea region until the is sparse, confined to a few records of Nothofagus early Miocene. leaves in boreholes (Hill, 1989; Cantrill, 2001). The The question is when did the Neogene vegetation lack of outcrop has hindered our understanding of the become extinct. Was it following the mid-Miocene patterns of changing diversity and the nature of the warm interval at c. 17 Ma when deglaciation might vegetation. This raises important questions as to have resulted in fjords extending from the Ross Sea to whether the cool temperate flora was rapidly replaced the heads of the valleys in the Transantarctic at this time or declined gradually throughout the Mountains, or was it following a mid-Pliocene warm Neogene. interval at about 3 Ma? A coupled GCM BIOME 4 For many years, it was believed that the model experiment using the USGS PRISM2 global Antarctic flora became extinct when the first ice data showed that the simulated Pliocene vegetation of sheets formed in the Oligocene. The huge volume the deglaciated coastal areas of Antarctica and over of glacially-derived shelf sediments made it difficult the Antarctic Peninsula would be cushion-forb-lichen- to envisage that a vegetation could have existed moss tundra, prostrate shrub tundra, and dwarf shrub anywhere on the continent. This notion conflicted or shrub tundra (Haywood et al., 2002). with evidence from palynology that suggested the presence of temperate forest well into the Oligocene in the Antarctic Peninsula region (Stuchlik, 1981) 8. Conclusions and East Antarctica (Kemp, 1975; Kemp and Barrett, 1975). However, radiometric dating of Based on numerous lines of evidence, we interpret sequences in the Antarctic Peninsula has proved the fossil plants of the MDF to represent a complex problematic and pollen assemblages indicate sub- mosaic of periglacial environments in which there was stantially older ages (Truswell, 1990). These abundant water from melting snow and glacier ice. assemblages are now known to be Eocene age. The diversity and dynamic nature of the environment Similar problems existed with the interpretation of resulted in a mosaic of tundra plant communities that palynological assemblages from East Antarctica included feldmark, cushion plant heath, prostrate where low yields and reworking made it difficult shrubland, localized turfs and mosses, and possibly to determine what part of the assemblage was in microbial mats. Ultimately, this vegetation became situ (Truswell, 1990). Consequently, the nature of extinct and was replaced by a polar desert as the ice 80 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 sheets expanded and the climate of the interior Askin, R.A., Ashworth, A.C., 1998. Palynomorphs from peat and became ultracold and ultraxeric. Future work to lacustrine beds in the Sirius Group of the central Transantarctic Mountains. Geological Society of America, Midwest Regional resolve the spatial and temporal aspects of this Meeting. ecosystem, together with improved identification of Askin, R.A., Markgraf, V., 1986. Palynomorphs from the Sirius the plants, will be important for developing a better Formation, Dominion Range, Antarctica. Antarc. J. U.S. 21 (5), understanding of the palaeoecology and biogeography 34–35. of Neogene polar plant communities. Askin, R.A., Raine, J.I., 2000. Oligocene and Early Miocene terrestrial palynology of the Cape Roberts Drillhole CRP-2/2A, Victoria Land Basin, Antarctica. Terra Antart. 7, 493–501. Barrett, P.J., Powell, R.D., 1982. Middle Cenozoic glacial beds at Acknowledgements Table Mountain, southern Victoria Land. In: Craddock, C. (Ed.), Antarctic Geoscience. University of Wisconsin Press, Madison, Collection of the samples on the Beardmore WI, pp. 1059–1067. Barrett, P.J., Adams, C.J., McIntosh, W.C., Swisher, C.C., Glacier would not have been possible without the Wilson, G.S., 1992. Geochronological evidence supports help of David Harwood, Mark Mabin, Jon DeVries, Antarctic deglaciation three million years ago. Nature 359, John Wright, and Marty Reed. 816–818. In the early stages of this study, David M. Moore, Barrett, P.J., Bleakley, N.L., Dickinson, W.W., Hannah, M.J., University of Reading, U.K., and Ole Bennike, Geo- Harper, M.A., 1997. Distribution of siliceous microfossils on Mount Feather, Antarctica, and the age of the Sirius Group. In: logical Survey of Denmark and Greenland, Ricci-Carlo, A. (Ed.), The Antarctic Region: Geological Kbbenhavn, provided helpful suggestions for the Evolution and Processes, Terra Antart., vol. 4, pp. 763–770. identification of the fossil achenes. We would also like Bliss, L.C., 1997. Arctic ecosystems of North America. In: to thank Dr. A. Newton, The Natural History Museum, Wielgolaski, F.E. (Ed.), Polar and Alpine Tundra, Ecosystems London, and Dr. L. Hedernas, Swedish Museum of of the World, vol. 3. Elsevier, Amsterdam, pp. 551–683. Brady, H.T., McKelvey, B.C., 1979. The interpretation of a tertiary Natural History, for discussions on the taxonomic tillite at Mount Feather, southern Victoria Land, Antarctica. placement of the mosses. Gary Comer provided the J. Glaciol. 22, 189–193. photograph of the fjord in southwestern Greenland. Brady, H.T., McKelvey, B.C., 1983. Some aspects of the Part of this work was supported while D.J.C. was Cenozoic glaciation of southern Victoria Land. J. Glaciol. employed at the British Antarctic Survey. The research 29, 343–349. Burckle, L.H., Potter Jr., N., 1996. Pliocene–Pleistocene diatoms in was supported by NSF/OPP Grants 9615252 and Paleozoic and Mesozoic sedimentary and igneous rocks from 0230696 to A.C.A. Antarctica: a Sirius problem solved. Geology 24, 235–238. Burckle, L.H., Stroeven, A.P., Bronge, C., Miller, U., Wasell, A., 1996. Deficiencies in the diatom evidence for a Pliocene References reduction of the east Antarctic ice sheet. Paleooceanography 11, 379–389. Ashworth, A.C., Kuschel, G., 2003. Fossil weevils (Coleoptera: Cantrill, D.J., 2001. Nothofagus from the Early Oligocene of Curculionidae) from latitude 85 degrees S Antarctica. Palae- Antarctica: implications for the vegetation history. Terra Antart. ogeogr. Palaeoclimatol. Palaeoecol. 191, 191–202. 8, 401–406. Ashworth, A.C., Preece, R.C., 2003. The first freshwater molluscs Cantrill, D.J., Poole, I., 2002. Cretaceous to Tertiary patterns of from Antarctica. J. Molluscan Stud. 69, 89–92. diversity change in the Antarctic Peninsula. In: Owens, A., Ashworth, A.C., Thompson, F.C., 2003. Palaeontology: a fly in the Crame, J.A. (Eds.), Palaeobiogeography and Biodiversity biogeographic ointment. Nature 423, 135–136. Change: A Comparison of the Ordovician and Mesozoic– Ashworth, A.C, Harwood, D.M., Webb, P.-N., Mabin, M.C.G., Cenozoic Radiations, Spec. Publ. -Geol. Soc. London, vol. 194, 1997. A weevil from the heart of Antarctica. In: Ashworth, pp. 141–152. A.C., Buckland, P., Sadler, J. (Eds.), Studies in Quaternary Carlquist, S., 1987. Pliocene Nothofagus wood from the Trans- Entomology, Quaternary Proceedings, vol. 5. Wiley, Chichester, antarctic Mountains. Aliso 11, 571–583. UK, pp. 15–21. Chapin III, F.S., Shaver, G.R., 1985. Arctic. In: Chabot, B.F., Askin, R.A., 1992. Late Cretaceous–early Tertiary Antarctic Mooney, H.A. (Eds.), Physiological Ecology of North outcrop evidence for past vegetation and climates. In: American Plant Communities. Chapman & Hall, London, Kennett, J.P., Warnke, D.A. (Eds.), The Antarctic Palaeoenvir- pp. 16–40. onment: A Perspective on Global Change, Antarctic Research Cheng, Fu-Chih, 2003. Classification of closed contours. Ph.D. Series, vol. 56. American Geophysical Union, Washington, pp. Thesis. North Dakota State University, Fargo, North Dakota, 61–73. U.S.A. A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82 81

Convey, P., 1996. Overwintering strategies of terrestrial inverte- In: Beyer, L., Bflter, M. (Eds.), Geoecology of Antarctic Ice- brates in Antarctica—the significance of flexibility in extremely Free Coastal Landscapes, Ecological Studies, vol. 154. Springer- seasonal environments. Eur. J. Entomol. 93, 489–505. Verlag, Berlin, pp. 327–374. Cronquist, A., 1981. An Integrated System of Classification of Kellogg, D.E., Kellogg, D.B., 1996. Diatoms in south pole ice: Flowering Plants. Columbia Univ. Press, New York. 1262 pp. implications for eolian contamination of Sirius Group deposits. DeConto, R.M., Pollard, D., 2003. Rapid Cenozoic glaciation of Geology 24, 115–118.

Antarctica induced by declining atmospheric CO2. Nature 421, Kemp, E.M., 1975. Palynology of Leg 28 drillsite, Deep Sea 245–249. Drilling Project. In: Hayes, D.E., Frakes, L.S. (Eds.), Initial Denton, G.H., Prentice, M.L., Kellogg, D.E., Kellogg, T.B., 1984. Reports of the Deep Sea Drilling Project, vol. 28. United States Late Tertiary history of the Antarctic ice sheet: evidence from Government Printing Office, Washington DC, pp. 599–621. the Dry Valleys. Geology 12, 263–267. Kemp, E.M., Barrett, P.J., 1975. Antarctic glaciation and early Dettmann, M.E., Thomson, M.R.A., 1987. Cretaceous palyno- Tertiary vegetation. Nature 258, 507–508. morphs from the James Ross Island Area, Antarctica—a pilot Lawver, L.A., Gahagan, L.M., Coffin, M.F., 1992. The development study. Br. Antarct. Surv. Bull. 77, 13–59. of paleoseaways around Antarctica. In: Kennett, J.P., Warnke, Dettmann, M.E., Pocknall, D.T., Romero, E.J., Zamaloa, M.C., D.A. (Eds.), The Antarctic Paleoenvironment: A Perspective on 1990. Nothofagidites Erdtman ex Potonie´, 1960; a catalogue of Global Change, Antarctic Research Series, vol. 56. American species with notes on the paleogeographic distribution of Geophysical Union, pp. 7–30. Nothofagus Bl. (Southern Beech). N.Z. Geol. Surv. Paleontol. Lewis Smith, R.I., 1994. Vascular plants as bioindicators of regional Bull. 60, 1–79. warming in Antarctica. Oecologia 99, 322–328. Fleming, R.F., Barron, J.A., 1996. Evidence of Pliocene Nothofagus Lockhart, P.J., McLenachan, P.A., Havell, D., Glenny, D., Huson, in Antarctica from Pliocene marine sedimentary deposits (DSDP D., Jenson, U., 2001. Phylogeny, radiation, and transoceanic Site 274). Mar. Micropalaeontol. 27, 227–236. dispersal of new Zealand alpine buttercups: molecular evidence Francis, J.E., Hill, R.S., 1996. Fossil plants from the Pliocene Sirius under split decomposition. Ann. Mo. Bot. Gard. 88, 458–477. Group, Transantarctic Mountains: evidence for climate from Marchant, D.R., Lewis, A.R., Phillips, W.M., Moore, E.J., Souchez, growth rings and fossil leaves. Palaios 11, 389–396. R.A., Denton, G.H., Sugden, D.E., Potter Jr., N., Landis, G.P., Gersonde, R., Kyte, F.T., Bleil, V., Diekman, B., Flores, J.A., Gehl, 2002. Formation of patterned ground and sublimation till over K., Grahl, G., Hagen, R., Kohn, G., Sierro, F.J., Voler, D., Miocene glacier ice in Beacon Valley, southern Victoria Land, Abelmann, A., Bostwick, J.A., 1997. Geological record and Antarctica. GSA Bull. 114, 718–730. reconstruction of the late Pliocene impact of the Eltanin asteroid Mayewski, P.A., 1975. Glacial geology and the late Cenozoic in the Southern Ocean. Nature 390, 357–363. history of the Transantarctic Mountains, Antarctica. Ohio State Hambrey, M.J., Webb, P.-N., Harwood, D.M., Krissek, L.A., 2003. University, Institute of Polar Studies Report No. 56, Columbus, Neogene glacial record from the Sirius Group of the Shackleton Ohio. 168 p. Glacier region, central Transantarctic Mountains, Antarctica. Mayewski, P.A., Goldthwaite, R.P., 1985. Glacial events in the Geol. Soc. Amer. Bull. 115, 994–1015. Transantarctic Mountains: a record of the East Antarctic Ice Harwood, D.M., 1986. Recycled marine microfossils from basal Sheet. In: Turner, M.D., Splettstoesser, J.F. (Eds.), Geology of debris-ice in ice-free valleys of southern Victoria Land, the Central Transantarctic Mountains, Antarctic Research Antarctic. Antarc. J. U.S. 21, 101–103. Series, vol. 36. American Geophysical Union, Washington Haywood, A.M., Francis, J.E., Sellwood, B.W., 2002. Global DC, pp. 275–324. middle Pliocene biome reconstruction: a data/model synthesis. McKelvey, B.C., Webb, P.-N., Harwood, D.M., Mabin, M.C.G., Geochem. Geophys. Geosystems 3, 1–18. 1991. The Dominion Range Sirius Group—a record of late Hill, R.S., 1989. Fossil leaf, Antarctic Cenozoic history from the Pliocene–early Pleistocene Beardmore Glacier. In: Thomson, CIROS-1 Drillhole, McMurdo Sound. N.Z. Dep. Sci. Ind. Res. M.R.A, Crame, J.A., Thomson, J.W. (Eds.), Geological Bull. 245, 143–144. Evolution of Antarctica. Cambridge Univ. Press, Cambridge, Hill, R.S., Truswell, E.M., 1993. Nothofagus fossils in the Sirius pp. 599–607. Group Transantarctic Mountains. In: Kennett, J.P., Warnke, D. Mercer, J.H., 1972. Some observations on the glacial geology of the (Eds.), The Antarctic Paleoenvironment: A Perspective on Beardmore Glacier area. In: Adie, R.J. (Ed.), Antarctic Geology Global Change, Antarctic Research Series, vol. 60. American and Geophysics. Universitetsforlaget, Oslo, pp. 427–433. Geophysical Union, WA, pp. 67–73. Mildenhall, D.C., 1989. Terrestrial palynology. In: Barrett, P.J. Hill, R.S., Jordan, G.J., 1996. Macrofossils as indicators of Plio- (Ed.), Antarctic Cenozoic History from the CIROS-1 Drillhole, Pleistocene climates in Tasmania. Pap. & Proc. Roy. Soc. McMurdo Sound, New Zealand Department of Scientific and Tasmania 130, 9–15. Industrial Research Bulletin, vol. 245, pp. 119–247. Hill, R.S., Harwood, D.M., Webb, P.N., 1996. Nothofagus beard- Moore, D.M., 1983. Flora of Tierra del Fuego. Anthony Nelson, morensis (Nothofagaceae), a new species based on leaves from England. the Pliocene Sirius Group, Transantarctic Mountains, Antarctica. Passchier, S., 2001. Provenance of the Sirius Group and related Rev. Palaeobot. Palynol. 94, 11–24. Upper Cenozoic glacial deposits from the Transantarctic Kappen, L., Schroeter, B., 2002. 18 plants and lichens in the Mountains, Antarctica: relation to landscape evolution and ice Antarctic, their way of life and their relevance to soil formation. sheet drainage. Sediment. Geol. 144, 263–290. 82 A.C. Ashworth, D.J. Cantrill / Palaeogeography, Palaeoclimatology, Palaeoecology 213 (2004) 65–82

Pole, M., Hill, R.S., Harwood, D., 2000. Eocene plant macrofossils Transantarctic Mountains rift flank constrained by cosmogenic from erratics, McMurdo Sound, Antarctica. In: Stilwell, J.D., isotope measurements. Glob. Planet. Change 23, 145–172. Feldmann, R.M. (Eds.), Palaeobiology and Palaeoenvironments Webb, P.-N., Harwood, D.M., 1987. Terrestrial flora of the Sirius of Eocene Fossiliferous Erratics, McMurdo Sound, East formation: its significance for Late Cenozoic glacial history. Antarctica, Antarctic Research Series, vol. 76. American Geo- Antarct. J. U.S. 22, 7–11. physical Union, Washington, pp. 243–251. Webb, P.-N., Harwood, D.M., 1993. Pliocene Nothofagus (southern Poole, I., 2002. Systematics of Cretaceous and Tertiary Nothofa- beech) from Antarctica: phytogeography, dispersal strategies, goxylon: implications for Southern Hemisphere biogeography and survival in high latitude glacial deglacial environments. In: and evolution of Nothofagaceae. Aust. Syst. Bot. 15, 247–276. Alden, J., Mastrantonio, J.C., Odum, S. (Eds.), Forest Develop- Raine, J.I., 1998. Terrestrial palynomorphs from Cape Roberts ment in Cold Climates. Plenum, New York, pp. 135–165. Project Drillhole CRP-1, Ross Sea, Antarctica. Terra Antart. 5, Webb, P.N., Harwood, D.M., McKelvey, B.C., Mercer, J.H., Stott, 539–548. L.D., 1984. Cenozoic marine sedimentation and ice volume Raine, I.J., Askin, R.A., 2001. Vegetation history and climate from variation on the East Antarctic craton. Geology 12, 287–291. the early Oligocene to the early Miocene, Cape Roberts, Ross Webb, P.N., McKelvey, B.C., Harwood, D.M., Mabin, M.C.G., Sea Region, Antarctica. Terra Antart. 8, 389–400. Mercer, J.H., 1987. Sirius formation of the Beardmore Glacier Retallack, G.J., Krull, E.S., Bockheim, J.G., 2001. New grounds for region. Antarc. J. U.S. 21, 8–13. reassessing palaeoclimate of the Sirius Group, Antarctica. Webb, P.N., Harwood, D.M., Mabin, M.C.G., McKelvey, B.C., J. Geol. Soc. (Lond.) 158, 925–935. 1996. A marine and terrestrial Sirius Group succession, middle Stroeven, A.P., Prentice, M.L., 1997. A case for Sirius Group alpine Beardmore Glacier–Queen Alexandra Range, Transantarctic glaciation at Mount Fleming, South Victoria Land, Antarctica: a Mountains, Antarctica. Mar. Micropalaeontol. 27, 273–297. case against Pliocene East Antarctic ice sheet reduction. Geol. Wilson, G.S., Harwood, D.M., Askin, R.A., Levy, R.H., 1998. Late Soc. Amer. Bull. 109, 825–840. Neogene Sirius Group strata in Reedy Valley, Antarctica: a Stroeven, A.P., Prentice, M.L., Kleman, J., 1996. On marine multiple-resolution record of climate, ice-sheet and sea level microfossil transport and pathways in Antarctica during the late events. J. Glaciol. 44, 437–447. Neogene: evidence from the Sirius Group at Mount Fleming. Wilson, G.S., Barron, J.A., Ashworth, A.C., Askin, R.A., Carter, Geology 24, 727–730. J.A., Curren, M.G., Dalhuisen, D.H., Friedman, E.I., Fyodorov- Stuchlik, L., 1981. Tertiary pollen spectra from the Ezcurra Inlet Davidov, D.G., Gilichinsky, D.A., Harper, M.A., Harwood, Group, Admiralty Bay, King George Island (South Shetland D.M., Hiemstra, J.F., Janecek, T.H., Licht, K.J., Ostroumov, Islands, Antarctica). Stud. Geol. Pol. 72, 109–132. V.E., Powell, R.D., Rivkina, E.M., Rose, S.A., Stroeven, A.P., Truswell, E.M., 1983. Recycled Cretaceous and Tertiary pollen and Stroeven, P., van der Meer, J.J.M., Wizevich, M.C., 2002. The spores in Antarctic marine sediments: a catalogue. Palae- Mount Feather Diamicton of the Sirius Group: an accumulation ontographica. B 186, 21–174. of indicators of Neogene Antarctic glacial and climatic history. Truswell, E.M., 1990. Cretaceous and Tertiary vegetation of Palaeogeog., Palaeoclimatol., Palaeoecol. 28, 1–15. Antarctica: a palynological perspective. In: Taylor, T.N., Taylor, Winkworth, R.C., Wagstaff, S.J., Glenny, D., Lockhart, P.J., 2002. E.L. (Eds.), Antarctic Paleobiology. Springer-Verlag, New York, Plant dispersal N.E.W.S. from New Zealand. Trends Ecol. Evol. pp. 71–88. 17, 514–520. Van der Wateren, F.M., Dunai, T.J., van Balen, R.T., Klas, W., Zachos, J.C., Pegani, M., Stone, L., Thomas, E., Billups, K., 2001. Verbers, A.L.L.M., Passchier, S., Herpers, U., 1999. Contrasting Trends, rhythms, and aberrations in global climates 65 Ma to denudation histories of different structural regions in the present. Science 292, 293–686.